Search Results

Clarius Median Nerve AI is intended for segmentation and semi-automatic non-invasive measurements of the median nerve cross-sectional area on ultrasound data acquired by the Clarius Ultrasound Scanner (i.e., linear array scanners). The user shall be a healthcare professional trained and qualified in ultrasound. The user retains the responsibility of confirming the validity of the measurements based on standard practices and clinical judgment. Clarius Median Nerve Al is indicated for use in adult patients only.

Clarius Median Nerve AI is a machine learning algorithm that is integrated into the Clarius App software as part of the complete Clarius Ultrasound Scanner system for use in musculoskeletal ultrasound applications, specifically intended for segmentation and measurement of the cross-sectional area of the median nerve. Clarius Median Nerve AI is intended for use by trained healthcare practitioners for measurement of the cross-sectional area (CSA) of the median nerve on ultrasound data acquired by the Clarius Ultrasound Scanner system (i.e., linear array scanners) using a deep learning image segmentation algorithm.

During the ultrasound imaging procedure, the anatomical site is selected through a preset software selection (i.e., Hand/Wrist) from the Clarius App in which Clarius Median Nerve AI will segment the median nerve in transverse view (with a segmentation mask placed on the ultrasound image) and engage to automatically place calipers on the segmentation mask to measure the median nerve's cross-sectional area.

Clarius Median Nerve AI operates by performing the following tasks:
• Automatic detection and measurement of the median nerve in transverse view

Clarius Median Nerve AI operates by identifying and segmenting the median nerve in the forearm and wrist and performs automatic measurements of the median nerve's cross-sectional area. The user has the option to manually adjust the measurements made by Clarius Median Nerve AI by moving the caliper crosshairs. Clarius Median Nerve AI does not perform any functions that could not be accomplished manually by a trained and qualified user.

Clarius Median Nerve AI is an assistive tool intended to inform clinical management and is not intended to replace clinical decision-making. The clinician retains the ultimate responsibility of ascertaining the measurements based on standard practices and clinical judgment. Clarius Median Nerve AI is indicated for use in adult patients only.

Clarius Median Nerve AI is integrated into the Clarius App software, which is compatible with iOS and Android operating systems two versions prior to the latest iOS or Android stable release build and is intended for use with the following Clarius Ultrasound Scanner system transducers (previously 510(k)-cleared in K213436). Clarius Median Nerve AI is not a stand-alone software device.

Here's an analysis of the acceptance criteria and the study proving the device meets them, based on the provided FDA 510(k) clearance document for Clarius Median Nerve AI:

Acceptance Criteria and Device Performance

1. A table of acceptance criteria and the reported device performance:

2. Sample size used for the test set and the data provenance:

• Test Set Sample Size: 182 images collected from 126 subjects. Some subjects had images collected at both forearm and wrist levels, accounting for the image count exceeding subject count.
• Data Provenance: Retrospective analysis of anonymized ultrasound images obtained from a multi-center database.
  • Countries of Origin: United States (majority - 130 images), Canada, Brazil, United Kingdom, Australia, Belgium, Germany, South Africa, Dominican Republic, Poland, The Netherlands, and Philippines.
  • Retrospective/Prospective: Retrospective. Data was previously collected and stored on a cloud platform.

3. Number of experts used to establish the ground truth for the test set and the qualifications of those experts:

• Number of Experts: 3 expert reviewers.
• Qualifications of Experts: Qualified experts with relevant (i.e., musculoskeletal) ultrasound experience. Specific details on years of experience or exact specializations (e.g., radiologist, sonographer, etc.) are not provided in the document, but it states they were "experienced ultrasound reviewers/clinicians."

4. Adjudication method for the test set:

• Adjudication Method: "To aggregate measurements from different truthers, the mean of the three values was taken and was treated as one reviewer mean." This suggests a form of consensus ground truth based on averaging individual expert measurements.
• Each reviewer was blinded to the Clarius Median Nerve AI output and the other reviewers' annotations.

5. If a multi-reader multi-case (MRMC) comparative effectiveness study was done, If so, what was the effect size of how much human readers improve with AI vs without AI assistance:

• The study described is a standalone (algorithm only) performance evaluation against human expert measurements, not a multi-reader multi-case (MRMC) comparative effectiveness study assessing human reader improvement with AI assistance.
• Therefore, no effect size for human reader improvement with AI assistance is provided or applicable from this document.

6. If a standalone (i.e. algorithm only without human-in-the-loop performance) was done:

• Yes, a standalone performance evaluation was done. The "Clinical Performance Evaluation Summary" and "Summary of the Clinical Verification Study" describe the AI model's measurements being compared directly against manual measurements from human experts without the experts using the AI as an assistive tool during their measurement process. The experts were explicitly "blinded to the Clarius Median Nerve AI output."

7. The type of ground truth used:

• Expert Consensus / Expert Manual Measurement: The ground truth for the test set measurements was established by manual measurements performed individually by three qualified human experts, and then aggregated by taking the mean of their three values.

8. The sample size for the training set:

• The document states that the Clarius Median Nerve AI Deep Neural Network (DNN) model was developed and trained using three data sets: training, tuning (validation), and internal testing.
• However, the exact sample size for the training set is NOT explicitly stated in the provided document. It mentions that data for model development was "collected from the Clarius Cloud and/or partner clinics" and partitioned, but it doesn't quantify the size of the training partition.

9. How the ground truth for the training set was established:

• The document states that the "internal test data was fully independent of the training/tuning dataset and was labelled by experts."
• By inference, the training and tuning (validation) data sets would also have had their ground truth established by experts' labeling, similar to the internal test set. However, the specific method (e.g., number of experts, adjudication) for the training data's ground truth is not detailed, only that it was "labelled by experts."

Ask a specific question about this device